Exploring temporally smooth representations via sparse coding and its biological plausibility in the human brain's visual system.
We consider the relationship between representations of natural images in a temporally smooth sequence (i.e. consecutive frames in a video). Traditionally, sparse coding methods learn representations of images in isolation. Here, we learn an image's sparse representation with the previous image's representation as a starting point. Our investigation links neuroscience and representation learning and builds on the increasingly popular field of machine learning, specifically neuro-biologically inspired learning models.